Prepare FishGlob Dataset

Script 1 for Kitchel et al. 2023 in prep taxonomic diversity manuscript.

library(tidyverse)

Registered S3 methods overwritten by 'dbplyr':
  method         from
  print.tbl_lazy     
  print.tbl_sql      
── Attaching packages ───────────────────────────────────────────────── tidyverse 1.3.2 ──✔ tibble  3.1.8      ✔ purrr   0.3.5 
✔ tidyr   1.2.1      ✔ dplyr   1.0.10
✔ readr   2.1.3      ✔ forcats 0.5.2 ── Conflicts ──────────────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::between()   masks data.table::between()
✖ dplyr::collapse()  masks nlme::collapse()
✖ tidyr::expand()    masks Matrix::expand()
✖ dplyr::filter()    masks stats::filter()
✖ dplyr::first()     masks data.table::first()
✖ dplyr::lag()       masks stats::lag()
✖ dplyr::last()      masks data.table::last()
✖ tidyr::pack()      masks Matrix::pack()
✖ purrr::transpose() masks data.table::transpose()
✖ tidyr::unpack()    masks Matrix::unpack()

library(sp)
library(raster)

Attaching package: ‘raster’

The following object is masked from ‘package:dplyr’:

    select

The following object is masked from ‘package:lme4’:

    getData

The following object is masked from ‘package:nlme’:

    getData

library(rgeos)

rgeos version: 0.5-9, (SVN revision 684)
 GEOS runtime version: 3.10.2-CAPI-1.16.0 
 Please note that rgeos will be retired by the end of 2023,
plan transition to sf functions using GEOS at your earliest convenience.
 GEOS using OverlayNG
 Linking to sp version: 1.4-7 
 Polygon checking: TRUE 

library(rgbif)
library(viridis)

Loading required package: viridisLite

library(gridExtra)

Attaching package: ‘gridExtra’

The following object is masked from ‘package:dplyr’:

    combine

library(rasterVis)
library(concaveman)
library(sf)
library(viridis)
library(cowplot)
library(data.table)
set.seed(1)

Pull in compiled and cleaned data from FishGlob, downloaded on November 28, 2022 (V 1.5). This is typically compiled by Dr. Aurore Maureaud. This includes public and private data, and therefore link cannot be shared. However, with editing, you can run analyses for public trawl surveys.

FishGlob_1.5 <- fread(here::here("data","FISHGLOB_v1.5_clean.csv"))

|--------------------------------------------------|
|==================================================|
|--------------------------------------------------|
|==================================================|

This version of FishGlob leaves out seasons for GMEX, fix here

#add season to GMEX to survey unit

FishGlob_1.5[survey == "GMEX", survey_unit := paste0(survey,"-",season)]

Also adding in seasons for NIGFS

#add season to GMEX to survey unit

FishGlob_1.5[survey == "NIGFS", survey_unit := paste0(survey,"-",quarter)]

Also add seasons for Nor-BTS

FishGlob_1.5[survey == "Nor-BTS" & month %in% c(1:6), survey_unit := "Nor-BTS-1"][survey == "Nor-BTS" & month %in% c(7:12), survey_unit := "Nor-BTS-3"]

ZAF (South Africa) has distinct Atlantic and Indian surveys (split at ~20.01˚ E, Cape Agulhas)

FishGlob_1.5[survey == "ZAF" & longitude <20.01, survey_unit := "ZAF-ATL"][survey == "ZAF" & longitude >= 20.01, survey_unit := "ZAF-IND"]

Region names

sort(unique(FishGlob_1.5[,survey_unit]))

 [1] "AI"          "BITS-1"      "BITS-4"      "CHL"         "COL"         "DFO-HS"     
 [7] "DFO-NF"      "DFO-QCS"     "DFO-SOG"     "DFO-WCHG"    "DFO-WCVI"    "EBS"        
[13] "EVHOE"       "FALK"        "FR-CGFS"     "GIN"         "GMEX-Fall"   "GMEX-Summer"
[19] "GOA"         "GRL-DE"      "GSL-N"       "GSL-S"       "ICE-GFS"     "IE-IGFS"    
[25] "IS-MOAG"     "IS-TAU"      "MEDITS"      "MRT"         "NAM"         "NEUS-Fall"  
[31] "NEUS-Spring" "NIGFS-1"     "NIGFS-4"     "Nor-BTS-1"   "Nor-BTS-3"   "NS-IBTS-1"  
[37] "NS-IBTS-3"   "NZ-CHAT"     "NZ-ECSI"     "NZ-SUBA"     "NZ-WCSI"     "PT-IBTS"    
[43] "ROCKALL"     "S-GEORG"     "SCS-FALL"    "SCS-SPRING"  "SCS-SUMMER"  "SEUS-fall"  
[49] "SEUS-spring" "SEUS-summer" "SWC-IBTS-1"  "SWC-IBTS-4"  "WBLS"        "WCANN"      
[55] "WCTRI"       "ZAF"         "ZAF-ATL"     "ZAF-IND"    

##Data Replacements ####Greenland (version in FishGlob 1.5 is missing lengths and therefore biomass values) This version was obtained directly from Karl-Michael Werner karl-michael.werner@thuenen.de who now manages the Greenland survey September 2023. He is based in Germany.

greenland <- 

####Norway Prepped by Laurene Pecuchet (U Trömso, Norway) September 2023 to replace what’s in FishGlob 1.5 because IMR “are quite concerned that FishGlob, and other studies, have been using a”flawed” multi-surveys dataset that is available in NMDC (data portal of IMR). Turns out that this dataset was put publicly by miscommunication on NMDC after one published paper in Scientific Reports, and I think they only realized the existence of this dataset just the last year as some papers are coming out using it (especially the one from Cesc Gordo-Vilaseca in PNAS https://www.pnas.org/doi/10.1073/pnas.2120869120). They are now trying to make some damage controls to make sure that this dataset is not used ever again in the future, but that cleanded and standardised datasets of the Barents Sea survey that are publicly available in NMDC are used instead of.

September 14: From Laurene, “I send you in attachment the “new” IMR survey formatted for Fishglob. I have done some small check of the dataset, and so far everything looks good, but I didn’t do a deep check yet, but I don’t see why there should be any problems with it….For your study, I think it is also important that you know that there has been some inconsistencies in taxonomic descriptions in the Barents Sea so that some species should be considered at the genus level instead of for biodiversity analysis, I send you in attach an excel (Barents Sea Fish Reference List.csv) file that summarize which species might be a misidentification and which one should be considered and merged.” All of these files now live in “data/Norway_Sep2023”

Given Barents Sea Fish Reference List.csv, there are some rare species that should only be trusted as a true ID if there are particular people on the boat. For consistency, these species will be removed. Despite the following information, these species will be maintained. -Current disagreement about whether Lycodes rossi and Lycodes reticulatus are different or same, will keep separate for now. -Specimens of Anarhichas less than 10cm are often misidentified, but we will still keep three distinct species

#to fix naming, link species names in data to species names in barents sea spp fix
norway_clean_spp <- unique(norway_clean[,.(accepted_name, verbatim_name)])

Error in h(simpleError(msg, call)) : 
  error in evaluating the argument 'x' in selecting a method for function 'unique': could not find function "."

Delete Greenland and Norway

FishGlob_1.5 <- FishGlob_1.5[!(survey %in% c("Nor-BTS", "GRL-DE"))]

Add in updated Greenland and Norway data

FishGlob_1.5 <-rbind(FishGlob_1.5,norway)
FishGlob_1.5 <-rbind(FishGlob_1.5,greenland)

##Preliminary Data Cuts ###Specific Regional Changes Before Cutting to 10 years only

GSL - North: we have data 1980-2019, but gear changes in 2004/2005, so let’s use later portion (more consistent months of sampling; 2005-2019; 15 years) - South: we have data 1970-2019, but gear/vessel changes in 1985 and again in 1992, so again let’s use later portion (1992-2019; 27 years) - See this github issue

#identify haul_ids of hauls we should remove from GSL surveys
haul_ids_to_remove_GSL <- unique(FishGlob_1.5[(survey == "GSL-N" & year < 2005)|(survey == "GSL-S" & year < 1992),haul_id])

FishGlob_1.5 <- FishGlob_1.5[!(haul_id %in% haul_ids_to_remove_GSL),] #remove hauls before consistent gear/vessel was used

Nor-BTS
- Overlap between IBTS and Nor-BTS surveys below 62˚latitude, so delete all hauls that occur below 62˚latitude

#identify haul_ids of hauls we should remove from Nor-BTS surveys
haul_ids_to_remove_Nor_BTS <- unique(FishGlob_1.5[(survey == "Nor-BTS" & latitude  < 62),haul_id])

FishGlob_1.5 <- FishGlob_1.5[!(haul_id %in% haul_ids_to_remove_Nor_BTS),] #remove hauls before consistent gear/vessel was used

SGEORG - From Martin Collins, “Most surveys were focussed on demersal fish on the South Georgia shelf (< 350 m), but surveys in 2003, 2010 and 2019 had some deeper trawls. The deeper trawls caught very different fish, so are unlikely to be of use to a long-term analysis, but I have left them in.”

-Delete all trawls below 350 M

###Because time is an essential component of these analyses, we will get rid of any survey x season combinations that are not sampled for at least 10 years

#new row for total number of years sampled
FishGlob_1.5[,years_sampled := length(unique(year)),.(survey_unit)]

summary(FishGlob_1.5$years_sampled) #ranges from 2 (DFO Straight of Georgia) to 57 (Northeast US)

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
   1.00   24.00   30.00   30.98   37.00   57.00 

#statistics about full dataset
nrow(FishGlob_1.5) 

[1] 4351504

length(unique(FishGlob_1.5[,survey])) 

[1] 45

length(unique(FishGlob_1.5[,survey_unit])) 

[1] 58

#remove observations for any regions x season combinations sampled less than 10 times
FishGlob.10year <- FishGlob_1.5[years_sampled >= 10,]

#statistics about reduced 10 year dataset
nrow(FishGlob.10year) 

[1] 4267620

length(unique(FishGlob.10year[,survey])) 

[1] 38

length(unique(FishGlob.10year[,as.character(survey_unit)])) 

[1] 49

#remove full database
rm(FishGlob_1.5)

###For taxonomic analyses, resolution to species is required. Therefore, we will exclude any observations not resolved to species.

FishGlob.10year.spp <- FishGlob.10year[rank %in% c("Species", "Subspecies"),] #4077730 total observations

#remove full species database
rm(FishGlob.10year)

#vector with all survey names
all_survey_units <- sort(unique(FishGlob.10year.spp[,survey_unit]))

#calculate # species per year
FishGlob.10year.spp_survey_year <- unique(FishGlob.10year.spp[,.(survey_unit, year, accepted_name)])

FishGlob.10year.spp_survey_year[,spp_count_survey_year := uniqueN(accepted_name),.(survey_unit, year)]

FishGlob.10year.spp_survey_year.r <-unique(FishGlob.10year.spp_survey_year[,.(survey_unit,  year, spp_count_survey_year)])

nrow(FishGlob.10year.spp_survey_year.r)

[1] 1257

#calculate # hauls per year
FishGlob.10year.spp_haulid_year <- unique(FishGlob.10year.spp[,.(survey_unit, year, haul_id)])

FishGlob.10year.spp_haulid_year[,haulid_count_survey_year := uniqueN(haul_id),.(survey_unit, year)]

FishGlob.10year.spp_haulid_year.r <-unique(FishGlob.10year.spp_haulid_year[,.(survey_unit,  year, haulid_count_survey_year)])

nrow(FishGlob.10year.spp_haulid_year.r)

[1] 1257

##Visually Inspect Distribution of Data Through Time and Space

##Spatial and Temporal Patterns in All Trawl Surveys

Let’s look at the number of hauls per year/month and year/quarter and year/season visually

#unique survey, survey_unit, year, month, quarter, season, haul_id, lat, lon
FishGlob.10year.uniquehauls <- unique(FishGlob.10year.spp[,.(survey, survey_unit, year,month,quarter,season,haul_id, latitude, longitude,haul_dur)])

#add column with adjusted longitude for few surveys that cross dateline (NZ-CHAT and AI)
FishGlob.10year.uniquehauls[,longitude_adj := ifelse((survey_unit %in% c("AI","NZ-CHAT") & longitude > 0),longitude-360,longitude)]

FishGlob.10year.uniquehauls[,haul_counts_per_survey_season_month :=uniqueN(haul_id),.(survey, month, season)][, #count # hauls per survey, season, and month
                     haul_counts_per_survey_quarter_month :=uniqueN(haul_id),.(survey, month, quarter)][,#count # hauls per survey, month, and quarter
                     total_hauls_survey :=uniqueN(haul_id),.(survey)][,#count # hauls per survey in all years
                                                        
              #proportion of hauls for each survey, season, and month divided by total # over all years
                     haul_proportion_survey_season :=haul_counts_per_survey_season_month/total_hauls_survey][,
              #proportion of hauls for each survey, quarter, and month divided by total # over all years
                     haul_proportion_survey_quarter :=haul_counts_per_survey_quarter_month/total_hauls_survey][,
                                                                                                               
                     haul_count_per_survey_year_month :=uniqueN(haul_id),.(year, survey_unit, month)][, #count # hauls per survey unit, year, and month
                     total_hauls_survey_year := uniqueN(haul_id),.(survey_unit,year)][, #count total # hauls per survey unit and year
                     #proportion of hauls for each survey unit and month divided by total # hauls within a survey unit within a year
                     haul_proportion_month_yearly := haul_count_per_survey_year_month/total_hauls_survey_year][, 

                     haul_count_per_survey_year_quarter :=uniqueN(haul_id),.(year, survey_unit, quarter)][, #count # hauls per survey unit, year, and month
                     #proportion of hauls for each survey unit and month divided by total # hauls within a survey unit within a year
                     haul_proportion_quarter_yearly := haul_count_per_survey_year_quarter/total_hauls_survey_year] 

FishGlob.10year.uniquehauls.season <- unique(FishGlob.10year.uniquehauls[,.(survey, survey_unit, month, season, haul_counts_per_survey_season_month,total_hauls_survey, haul_proportion_survey_season)]) #relative sampling by season across all years

FishGlob.10year.uniquehauls.quarter <- unique(FishGlob.10year.uniquehauls[,.(survey,survey_unit , month, quarter, haul_counts_per_survey_quarter_month,total_hauls_survey, haul_proportion_survey_quarter)]) #relative sampling by quarter across all years

FishGlob.10year.uniquehauls.annual.month <- unique(FishGlob.10year.uniquehauls[,.(survey, year, survey_unit, month, haul_count_per_survey_year_month,total_hauls_survey_year,haul_proportion_month_yearly)]) #relative sampling by month within years

FishGlob.10year.uniquehauls.annual.quarter <- unique(FishGlob.10year.uniquehauls[,.(survey, year, survey_unit, quarter, haul_count_per_survey_year_quarter,total_hauls_survey_year,haul_proportion_quarter_yearly)]) #relative sampling by month within years

#how does #hauls vary with season and month?
survey_season_month_hauls <- ggplot(FishGlob.10year.uniquehauls.season) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  facet_wrap(~survey,scales = "free_y") +
  theme_classic()

ggsave(survey_season_month_hauls, filename = "survey_season_month_hauls.pdf",path = here::here("figures","view_data"), height = 5, width = 15, units = "in")

#how does #hauls vary with quarter and month?
survey_quarter_month_hauls <- ggplot(FishGlob.10year.uniquehauls.quarter) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  facet_wrap(~survey,scales = "free_y") +
  theme_classic()

ggsave(survey_quarter_month_hauls, filename = "survey_quarter_month_hauls.pdf",path = here::here("figures","view_data"), height = 5, width = 15, units = "in")

#how does #hauls vary with year and month?
year_survey_month_hauls <- ggplot(FishGlob.10year.uniquehauls.annual.month) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  facet_wrap(~survey_unit,scales = "free_y") +
  theme_classic()

ggsave(year_survey_month_hauls, filename = "year_survey_month_hauls.pdf",path = here::here("figures","view_data"), height = 8, width = 16, units = "in")
ggsave(year_survey_month_hauls, filename = "year_survey_month_hauls.pdf",path = here::here("figures","view_data"), height = 8, width = 16, units = "in")

#how does #hauls vary with year and month?
year_survey_quarter_hauls <- ggplot(FishGlob.10year.uniquehauls.annual.quarter) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  facet_wrap(~survey_unit,scales = "free_y") +
  theme_classic()

ggsave(year_survey_quarter_hauls, filename = "year_survey_quarter_hauls.pdf",path = here::here("figures","view_data"), height = 8, width = 16, units = "in")
ggsave(year_survey_quarter_hauls, filename = "year_survey_quarter_hauls.pdf",path = here::here("figures","view_data"), height = 8, width = 16, units = "in")

Now, let’s look at how location of sampling varies by month of sampling and year of sampling

location_by_year <- ggplot(FishGlob.10year.uniquehauls) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  facet_wrap(~survey_unit, scales = "free") +
  theme_classic()

ggsave(location_by_year, filename = "location_by_year.pdf",path = here::here("figures","view_data"), height = 8, width = 12, units = "in")

ggsave(location_by_year, filename = "location_by_year.jpg",path = here::here("figures","view_data"), height = 8, width = 12, units = "in")

ggsave(location_by_year, filename = "location_by_year.eps",path = here::here("figures","view_data"), height = 8, width = 12, units = "in")

location_by_month <- ggplot(FishGlob.10year.uniquehauls) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  facet_wrap(~survey_unit, scales = "free") +
  theme_classic()

ggsave(location_by_month, filename = "location_by_month.pdf",path = here::here("figures","view_data"), height = 8, width = 12, units = "in")

ggsave(location_by_month, filename = "location_by_month.jpg",path = here::here("figures","view_data"), height = 8, width = 12, units = "in")

ggsave(location_by_month, filename = "location_by_month.eps",path = here::here("figures","view_data"), height = 8, width = 12, units = "in")

##Region Specific Data Processing

-Fredston et al. 2022 and Batt et al. 2017 informed North American data processing -Personal communication with Aurore Maureaud re: work by L. Pecuchet and R. Frelat and the supplementary material for Maureaud et al. 2019 informed European data processing -Additional data processing informed by data itself, and by FishGlob pdf summary documents -limit to max 3 months for each survey unit, representative of a ‘season’

####“AI”

ggplot(FishGlob.10year.uniquehauls.season[survey == "AI",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey == "AI",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey == "AI",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey == "AI",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey == "AI",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey == "AI",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "AI",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "AI",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

• Most hauls in 6,7,8
• Seemingly consistent spatial distribution through time
• No dramatic changes in spp richness

ai_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "AI" & month %in% c(6:8),haul_id])

####BITS (We have two surveys for BITS, quarter 1 and quarter 4) BITS 1

From Fredston et al. 2023, every year after 2000 has >400 hauls and most of the earlier years are <50

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "BITS-1",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "BITS-1",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "BITS-1",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "BITS-1",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "BITS-1",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "BITS-1",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "BITS-1",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "BITS-1",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep both months (2,3) -Seemingly consistent spatial distribution through time -Consistent # of species and # hauls after 2000

bits1_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "BITS-1" & month %in% c(2,3) & year > 2000,haul_id])

BITS4

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "BITS-4",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "BITS-4",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "BITS-4",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "BITS-4",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "BITS-4",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "BITS-4",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "BITS-4",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "BITS-4",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep (10,11,12) -Start in 2000 (starts in 1996, but gap in 1997 and 1998, and 1996 all in December; also spp richness in first survey very low; consistent # of hauls after 2000) -Seemingly consistent spatial distribution through time

bits4_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "BITS-4" & month %in% c(10:12) & year > 2000,haul_id])

####CHL (Chile)

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "CHL",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "CHL",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "CHL",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "CHL",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "CHL",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "CHL",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "CHL",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "CHL",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep (7,8,9) -Seemingly consistent spatial distribution through time -No major changes in spp richness through time -No major changes in # hauls through time

chl_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "CHL" & month %in% c(7:9),haul_id])

####DFO-NF

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "DFO-NF",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "DFO-NF",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "DFO-NF",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "DFO-NF",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "DFO-NF",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "DFO-NF",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "DFO-NF",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "DFO-NF",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep (10,11,12) -Seemingly consistent spatial distribution through time -No major changes in spp richness through time -No major changes in haulid through time

dfo_nf_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "DFO-NF" & month %in% c(10:12),haul_id])

####DFO-QCS

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "DFO-QCS",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "DFO-QCS",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "DFO-QCS",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "DFO-QCS",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "DFO-QCS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "DFO-QCS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "DFO-QCS",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "DFO-QCS",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep (7,8) -Seemingly consistent spatial distribution through time -No major changes in richness over time -No major changes in #hauls

dfo_qcs_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "DFO-QCS" & month %in% c(7,8),haul_id])

####EBS

-Sampling years prior to 1984 (data begin in 1982) were excluded from analysis due to large apparent increases in the number of species recorded in the first two years. (Batt et al. 2017)

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "EBS",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "EBS",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "EBS",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "EBS",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "EBS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "EBS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "EBS",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "EBS",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep (6,7,8) -Seemingly consistent spatial distribution through time -Per Batt et al. 2017, limit to >= 1984 -No clear changes in richness through time -No clear changes in # hauls through time

ebs_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "EBS" & month %in% c(6,7,8) & year >= 1984,haul_id])

####EVHOE

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "EVHOE",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "EVHOE",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "EVHOE",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "EVHOE",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "EVHOE",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "EVHOE",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "EVHOE",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "EVHOE",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep (10,11,12) -Seemingly consistent spatial distribution through time -Very low sampling in 2017 (and also low richness), exclude this year

evhoe_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "EVHOE" & month %in% c(10,11,12) & year != 2017 ,haul_id])

####FALK (excluded from final dataset)

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "FALK",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "FALK",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "FALK",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "FALK",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "FALK",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "FALK",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "FALK",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "FALK",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep February (2) only from 2004 onward (most consistent sampling) -Inconsistent spatial distribution through time, but this will be fixed in next step with spatial standardization

falk_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "FALK" & month %in% c(2) & year >= 2004, haul_id])

####FR-CGFS

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "FR-CGFS",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "FR-CGFS",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "FR-CGFS",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "FR-CGFS",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "FR-CGFS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "FR-CGFS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "FR-CGFS",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "FR-CGFS",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep 9,10,11 -Consistent spatial distribution through time -Seemingly consistent richness through time -Seeemingly consistent #hauls through time

fr_cgfs_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "FR-CGFS" & month %in% c(9,10,11), haul_id])

####GIN (excluded from final dataset)

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "GIN",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "GIN",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "GIN",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "GIN",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GIN",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GIN",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "GIN",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "GIN",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Exclude this region, no consistent sampling through time

gin_hauls_keep <- NULL

####GMEX -In the Gulf of Mexico, we restricted our analysis to data from 1984 - 2000 (full range 1982-2014); if all years had been used, the number of sites sampled in at least 85% of years would drop from 39 to 13. (Batt et al. 2017)

GMEX Fall

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "GMEX-Fall",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "GMEX-Fall",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "GMEX-Fall",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "GMEX-Fall",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GMEX-Fall",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GMEX-Fall",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "GMEX-Fall",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "GMEX-Fall",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep 9,10,11 -Inconsistent spatial distribution through time, will restrict to <-87.5 longitude -Seemingly consistent richness through time -Seeemingly consistent #hauls through time

gmex_fall_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "GMEX-Fall" & month %in% c(9,10,11) & longitude_adj < -87.5, haul_id])

GMEX Summer

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "GMEX-Summer",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "GMEX-Summer",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "GMEX-Summer",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "GMEX-Summer",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GMEX-Summer",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GMEX-Summer",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "GMEX-Summer",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "GMEX-Summer",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep months 5,6,7 -In consistent spatial distribution through time, but this will be fixed in spatial standardization step -Seemingly consistent richness before 2008 and 2008 onward through time -Seeemingly consistent #hauls through time -Jump from 2007 to 2008, when spatial footprint increases, so I will only use data from before 2008

gmex_summer_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "GMEX-Summer" & month %in% c(5,6,7) & year <2008, haul_id])

####GOA

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "GOA",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "GOA",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "GOA",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "GOA",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GOA",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GOA",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "GOA",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "GOA",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep months 6,7,8 -Consistent spatial distribution through time -Seemingly consistent richness -Seemingly consistent #hauls through time

goa_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "GOA" & month %in% c(6,7,8), haul_id])

####GRL-DE -From Beukhof et al. 2019, all surveys in October and November

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "GRL-DE",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "GRL-DE",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "GRL-DE",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "GRL-DE",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GRL-DE",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GRL-DE",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "GRL-DE",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "GRL-DE",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-No months in data set, but according to Beukhof et al. 2019, all sampling in October and November so keep all -Consistent spatial distribution through time -Seemingly consistent richness -# of hauls drops between 1991 and 1992, and both 1992 and 2017 so limit to years between (1993-2016)

grl_de_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "GRL-DE" & year %in% c(1993:2016), haul_id])

####GSL

GSL-N

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "GSL-N",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "GSL-N",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "GSL-N",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "GSL-N",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GSL-N",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GSL-N",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "GSL-N",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "GSL-N",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep 6,7,8 -Consistent spatial distribution through time -Seemingly consistent richness -# of hauls in 2005 is higher, so start in 2006

gsl_n_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "GSL-N" & year > 2005, haul_id])

GSL-S

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "GSL-S",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "GSL-S",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "GSL-S",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "GSL-S",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GSL-S",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "GSL-S",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "GSL-S",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "GSL-S",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep 8,9,10 -Consistent spatial distribution through time -Seemingly consistent richness -Seemingly consistent number of hauls

gsl_s_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "GSL-S" & month %in% c(8:10), haul_id])

####ICE-GFS

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "ICE-GFS",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "ICE-GFS",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "ICE-GFS",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "ICE-GFS",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "ICE-GFS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "ICE-GFS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "ICE-GFS",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "ICE-GFS",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep 2,3,4 -Consistent spatial distribution through time -Seemingly consistent richness -Seemingly consistent number of hauls

ice_gfs_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "ICE-GFS" & month %in% c(2:4), haul_id])

####IE-IGFS

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "IE-IGFS",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "IE-IGFS",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "IE-IGFS",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "IE-IGFS",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "IE-IGFS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "IE-IGFS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "IE-IGFS",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "IE-IGFS",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep 10,11,12 -Consistent spatial distribution through time after 2004 (sampled far east in 2003 and 2004) -Seemingly consistent richness -Seemingly consistent number of hauls

ie_igfs_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "IE-IGFS" & month %in% c(10:12) & year  > 2004, haul_id])

####IS-MOAG (excluded from final dataset)

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "IS-MOAG",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "IS-MOAG",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "IS-MOAG",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "IS-MOAG",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "IS-MOAG",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "IS-MOAG",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "IS-MOAG",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "IS-MOAG",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Sampling too scattered over time, excluding

is_moag_hauls_keep <- NULL

####MEDITS

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "MEDITS",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "MEDITS",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "MEDITS",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "MEDITS",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "MEDITS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "MEDITS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "MEDITS",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "MEDITS",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep all surveys in quarter 2 -Consistent spatial distribution through time -Seemingly consistent richness -Seemingly consistent number of hauls

medits_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "MEDITS", haul_id])

####MRT (excluded from final dataset)

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "MRT",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "MRT",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "MRT",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "MRT",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "MRT",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "MRT",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "MRT",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "MRT",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Sampling inconsistent, exclude completely

mrt_hauls_keep <- NULL

####NAM

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "NAM",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "NAM",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "NAM",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "NAM",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NAM",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NAM",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "NAM",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "NAM",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep surveys in 1 and 2 (most consistently sampled) -Consistent spatial distribution through time -Seemingly consistent richness except for 1998 (exclude) -Seemingly consistent number of hauls except for 1998 (exclude)

nam_hauls_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "NAM" & month %in% c(1,2) & year != 1998, haul_id])

####NEUS

NEUS Spring

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "NEUS-Spring",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "NEUS-Spring",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "NEUS-Spring",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "NEUS-Spring",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NEUS-Spring",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NEUS-Spring",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "NEUS-Spring",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "NEUS-Spring",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep 3,4,5 months -Inconsistent spatial distribution through time, but should be caught in standardization step below -Seemingly consistent richness (especially after 87, should be fixed with standardization step) -Seemingly consistent number of hauls (especially after 81, should be fixed with standardization step)

#calculate wgt_cpue (km^2 avg from sean Lucey) and wgt_h (all biomass values calibrated to standard pre 2009 30 minute tow)
FishGlob.10year.spp[survey == "NEUS", wgt_h := wgt/0.5][survey == "NEUS", wgt_cpue := wgt/0.0384][survey == "NEUS", num_h := num/0.5][survey == "NEUS", num_cpue := num/0.0384]


#also, for northeast, we are going to delete any hauls before 2009 that are outside of +/- 5 minutes of 30 minutes and 2009 forward that are outside of +/- 5 minutes of 20 minutes
neus_spring_keep <- unique(FishGlob.10year.uniquehauls[((survey_unit == "NEUS-Spring" & month %in% c(3:5) & year < 2009 & (haul_dur > 0.42 & haul_dur < 0.58)) |
                                                        (survey_unit == "NEUS-Spring" & month %in% c(3:5) & year >= 2009 & (haul_dur > 0.25  & haul_dur < 0.42))), haul_id])

NEUS Fall

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "NEUS-Fall",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "NEUS-Fall",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "NEUS-Fall",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "NEUS-Fall",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NEUS-Fall",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NEUS-Fall",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "NEUS-Fall",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "NEUS-Fall",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep 9,10,11 months -Inconsistent spatial distribution through time, but should be caught in standardization step below -Seemingly consistent richness (especially after 84, should be fixed with standardization step) -Seemingly consistent number of hauls (especially after 85, should be fixed with standardization step)

#also, for northeast, we are going to delete any hauls before 2009 that are outside of +/- 5 minutes of 30 minutes and 2009 forward that are outside of +/- 5 minutes of 20 minutes
neus_fall_keep <- unique(FishGlob.10year.uniquehauls[((survey_unit == "NEUS-Fall" & month %in% c(9,10,11) & year < 2009 & (haul_dur > 0.42 & haul_dur < 0.58)) |
                                                        (survey_unit == "NEUS-Fall" & month %in% c(9,10,11) & year >= 2009 & (haul_dur > 0.25  & haul_dur < 0.42))), haul_id])

####NIGFS Northern Ireland

Spring Northern Ireland (quarter 1)

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "NIGFS-1",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "NIGFS-1",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "NIGFS-1",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "NIGFS-1",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NIGFS-1",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NIGFS-1",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "NIGFS-1",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "NIGFS-1",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep 2,3,4 months -Inconsistent spatial distribution through time, but should be caught in standardization step below -Seemingly consistent richness -Seemingly consistent number of hauls

nigfs_1_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "NIGFS-1" & month %in% c(2,3,4), haul_id])

Spring Northern Ireland (quarter 1)

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "NIGFS-4",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "NIGFS-4",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "NIGFS-4",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "NIGFS-4",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NIGFS-4",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NIGFS-4",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "NIGFS-4",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "NIGFS-4",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Keep 10,11 months -Consistent spatial distribution through time, but should be caught in standardization step below -Seemingly consistent richness -Seemingly consistent number of hauls

nigfs_4_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "NIGFS-4" & month %in% c(10,11), haul_id])

####Nor-BTS

First half of year (1:6), Excluded from final dataset Nor-BTS-1

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "Nor-BTS-1",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "Nor-BTS-1",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "Nor-BTS-1",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "Nor-BTS-1",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "Nor-BTS-1"&month,]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "Nor-BTS-1"&month,]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "Nor-BTS-1",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "Nor-BTS-1",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use 1:3 months -Somewhat inconsistent spatial distribution through time (northern sites sampled later), but should be caught in standardization step below -Laurene Pecuchet (U Tromso) told us that only surveys 2004 and onwards work for biodiversity analyses -From 2004, consistent number of hauls except for 2013 which we will exclude

nor_bts_1_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "Nor-BTS-1" & month %in% c(1:3) & year >= 2004 & year != 2013, haul_id])

Second half of year (2:12) Nor-BTS-3

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "Nor-BTS-3",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "Nor-BTS-3",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "Nor-BTS-3",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "Nor-BTS-3",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "Nor-BTS-3",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "Nor-BTS-3",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "Nor-BTS-3",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "Nor-BTS-3",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 8,9,10 -Somewhat inconsistent spatial distribution through time, but should be caught in standardization step below -Number of hauls is variable, but no clear years to exclude -Laurene Pecuchet (U Tromso) told us that only surveys 2004 and onwards work for biodiversity analyses

nor_bts_3_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "Nor-BTS-3" & month %in% c(8:10) & year >= 2000, haul_id])

####NS-IBTS

NS-IBTS-1

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "NS-IBTS-1",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "NS-IBTS-1",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "NS-IBTS-1",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "NS-IBTS-1",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NS-IBTS-1",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NS-IBTS-1",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "NS-IBTS-1",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "NS-IBTS-1",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 1,2,3 -Consistent spatial distribution through time -Linear increase in richness, cutoff on # hauls more clear -Linear increase, but somewhat clear break between late 70s and mid-80s, only keep hauls after 1984

ns_ibts_1_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "NS-IBTS-1" & month %in% c(1:3) & year >= 1984, haul_id])

NS-IBTS-3

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "NS-IBTS-3",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "NS-IBTS-3",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "NS-IBTS-3",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "NS-IBTS-3",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NS-IBTS-3",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NS-IBTS-3",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "NS-IBTS-3",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "NS-IBTS-3",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 7,8,9 -Consistent spatial distribution through time -Consistent richness through time -Early years lower # hauls, will start at 1998

ns_ibts_3_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "NS-IBTS-3" & month %in% c(7:9) & year >= 1998, haul_id])

####NZ

NZ-CHAT

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "NZ-CHAT",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "NZ-CHAT",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "NZ-CHAT",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "NZ-CHAT",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NZ-CHAT",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NZ-CHAT",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "NZ-CHAT",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "NZ-CHAT",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 12,1,2 (NOTE THAT THIS NZ-CHAT SURVEY CROSSES YEAR, SO LUMP 1 and 2 with previous year) -Consistent spatial distribution through time -Seemingly consistent richness -Seemingly consistent number of hauls after 1995

nz_chat_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "NZ-CHAT" & month %in% c(12,1,2) & year >= 1995, haul_id])

NZ-ECSI

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "NZ-ECSI",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "NZ-ECSI",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "NZ-ECSI",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "NZ-ECSI",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NZ-ECSI",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NZ-ECSI",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "NZ-ECSI",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "NZ-ECSI",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 4,5,6 -Consistent spatial distribution through time -Seemingly consistent richness -Seemingly consistent number of hauls -Gap between 1995 and 2005, but we have 10 total years so we’ll keep for now

nz_ecsi_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "NZ-ECSI" & month %in% c(4,5,6), haul_id])

NZ-SUBA

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "NZ-SUBA",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "NZ-SUBA",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "NZ-SUBA",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "NZ-SUBA",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NZ-SUBA",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NZ-SUBA",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "NZ-SUBA",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "NZ-SUBA",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 11 and 12 -Consistent spatial distribution through time -Seemingly consistent richness -Far more hauls in 1990s, these early sampling years will be excluded (start in 2000)

nz_suba_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "NZ-SUBA" & month %in% c(11,12) & year >= 2000, haul_id])

NZ-WCSI

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "NZ-WCSI",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "NZ-WCSI",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "NZ-WCSI",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "NZ-WCSI",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NZ-WCSI",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "NZ-WCSI",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "NZ-WCSI",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "NZ-WCSI",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 3,4 -Consistent spatial distribution through time -Seemingly consistent richness -Linear decrease in # of hauls through time, leave out first two years with highest # hauls (>= 1995)

nz_wcsi_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "NZ-WCSI" & month %in% c(3,4) & year >= 1995, haul_id])

####PT-IBTS PT-IBTS

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "PT-IBTS",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "PT-IBTS",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "PT-IBTS",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "PT-IBTS",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "PT-IBTS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "PT-IBTS",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "PT-IBTS",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "PT-IBTS",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 9,10,11 -Consistent spatial distribution through time -Seemingly consistent richness -Seemingly consistent number of hauls

pt_ibts_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "PT-IBTS" & month %in% c(9,10,11), haul_id])

####ROCKALL

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "ROCKALL",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "ROCKALL",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "ROCKALL",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "ROCKALL",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "ROCKALL",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "ROCKALL",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "ROCKALL",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "ROCKALL",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 8,9 -Consistent spatial distribution through time -Seemingly consistent richness -Seemingly consistent number of hauls

rockall_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "ROCKALL" & month %in% c(8,9), haul_id])

####S-GEORG

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "S-GEORG",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "S-GEORG",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "S-GEORG",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "S-GEORG",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "S-GEORG",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "S-GEORG",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "S-GEORG",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "S-GEORG",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 1 and 2 -Consistent spatial distribution through time -Seemingly consistent richness except for 2003, will be excluded -Seemingly consistent number of hauls, except for 2012, will be excluded

s_georg_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "S-GEORG" & month %in% c(1,2) & !(year %in% c(2003,2012)), haul_id])

####SCS

Spring

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "SCS-SPRING",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "SCS-SPRING",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "SCS-SPRING",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "SCS-SPRING",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SCS-SPRING",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SCS-SPRING",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "SCS-SPRING",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "SCS-SPRING",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 2,3,4 -Inconsistent spatial distribution through time (northern latitudes only sampled in early years), only include longitudes < -62 and latitudes < 45.5 -Seemingly consistent richness -Number of hauls is variable, exclude super low and high numbers (1985,1994,2015,2019)

scs_spring_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "SCS-SPRING" & month %in% c(2,3,4) & !(year %in% c(1985,1994,2015,2019)) & longitude_adj < -62 & latitude < 45.5, haul_id])

SUMMER

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "SCS-SUMMER",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "SCS-SUMMER",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "SCS-SUMMER",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "SCS-SUMMER",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SCS-SUMMER",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SCS-SUMMER",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "SCS-SUMMER",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "SCS-SUMMER",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 6,7,8 -Consistent spatial distribution through time -Richness increases linearly, not a clear break point, using breakpoint from # of hauls, but will exclude 2010 which has a very high richness -# Hauls increases linearly from ~120 in 1970 to ~220 in 2020, not a clear breakpoint, but will go with 1986 because there is a jump between 85 and 86 -Gear change in 1983 (Ellingsen et al. 2015)

scs_summer_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "SCS-SUMMER" & month %in% c(6,7,8) & year >= 1986 & year != 2010, haul_id])

###SEUS

Spring

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "SEUS-spring",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "SEUS-spring",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "SEUS-spring",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "SEUS-spring",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SEUS-spring",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SEUS-spring",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "SEUS-spring",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "SEUS-spring",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 4,5,6 -Consistent spatial distribution through time -Consistent richness through time -# Hauls low in 1989 and 2018, will exclude

seus_spring_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "SEUS-spring" & month %in% c(4,5,6) & year != 1989 & year != 2018, haul_id])

Summer

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "SEUS-summer",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "SEUS-summer",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "SEUS-summer",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "SEUS-summer",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SEUS-summer",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SEUS-summer",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "SEUS-summer",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "SEUS-summer",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 7,8 -Consistent spatial distribution through time -Richness consistent through time -# Hauls low in first year, otherwise okay, just exclude first year (1989)

seus_summer_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "SEUS-summer" & month %in% c(7,8) & year != 1989, haul_id])

Fall

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "SEUS-fall",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "SEUS-fall",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "SEUS-fall",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "SEUS-fall",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SEUS-fall",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SEUS-fall",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "SEUS-fall",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "SEUS-fall",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 9,10,11 -Consistent spatial distribution through time -Richness consistent through time -# Hauls low in first year, otherwise okay, just exclude first year (1989)

seus_fall_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "SEUS-fall" & month %in% c(9,10,11) & year != 1989, haul_id])

####SWC-IBTS

Scotland Shelf Sea

SWC-IBTS 1

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "SWC-IBTS-1",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "SWC-IBTS-1",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "SWC-IBTS-1",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "SWC-IBTS-1",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SWC-IBTS-1",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SWC-IBTS-1",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "SWC-IBTS-1",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "SWC-IBTS-1",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 1,2,3 -Somewhat inconsistent spatial distribution through time, but this should be addressed in spatial standardization procedure -Richness consistent through time -# Hauls consistent except low in 1995, just exclude 1995

swc_ibts_1_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "SWC-IBTS-1" & month %in% c(1,2,3) & year != 1995, haul_id])

SWC-IBTS 4

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "SWC-IBTS-4",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "SWC-IBTS-4",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "SWC-IBTS-4",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "SWC-IBTS-4",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SWC-IBTS-4",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "SWC-IBTS-4",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "SWC-IBTS-4",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "SWC-IBTS-4",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Use months 10,11,12 -Somewhat inconsistent spatial distribution through time (southern latitudes only sampled in early years), but this should be addressed in spatial standardization procedure -Richness consistent through time (especially after mid 90s) -# Hauls consistent except low before 1995 and low in 2013, exclude these

swc_ibts_4_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "SWC-IBTS-4" & month %in% c(10,11,12) & year != 1995 & year >= 1995, haul_id])

####WCANN

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "WCANN",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "WCANN",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "WCANN",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "WCANN",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "WCANN",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "WCANN",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "WCANN",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "WCANN",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Here, one exception, will use four months (6,7,8,9) because all sampled consistently, and lower latitude areas sampled later in the summer consistently -Consistent spatial distribution through time -Richness consistent through time -# Hauls consistent through time

wcann_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "WCANN" & month %in% c(6:9), haul_id])

####WCTRI -Exclude because only 10 years and overlaps somewhat wiith WCANN

wctri_keep <- NULL

####ZAF

ATL

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "ZAF-ATL",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "ZAF-ATL",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "ZAF-ATL",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "ZAF-ATL",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "ZAF-ATL",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "ZAF-ATL",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "ZAF-ATL",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "ZAF-ATL",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Include 1,2,3 -Consistent spatial distribution through time -Richness consistent through time -# Hauls consistent through time after 1991

zaf_atl_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "ZAF-ATL" & month %in% c(1:3) & year >= 1991, haul_id])

IND

ggplot(FishGlob.10year.uniquehauls.season[survey_unit == "ZAF-IND",]) +
  geom_tile(aes(x = factor(month), y = factor(season), fill = haul_proportion_survey_season),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Season",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.quarter[survey_unit == "ZAF-IND",]) +
  geom_tile(aes(x = factor(month), y = factor(quarter), fill = haul_proportion_survey_quarter),color = "white") +
  scale_fill_viridis() +
  labs(x = "Month", y = "Quarter",fill = "Proportion of All Survey Hauls in FishGlob") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.month[survey_unit == "ZAF-IND",]) +
  geom_tile(aes(x = year, y = factor(month), fill = haul_proportion_month_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Month",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls.annual.quarter[survey_unit == "ZAF-IND",]) +
  geom_tile(aes(x = year, y = factor(quarter), fill = haul_proportion_quarter_yearly),color = "white") +
  scale_fill_viridis() +
  labs(x = "Year", y = "Quarter",fill = "Proportion of Annual Hauls") +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "ZAF-IND",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = year), size = 0.3, alpha = 0.5) +
  scale_color_viridis() +
  theme_classic()

ggplot(FishGlob.10year.uniquehauls[survey_unit == "ZAF-IND",]) +
  geom_point(aes(x = longitude_adj, y = latitude, color = month), size = 0.3, alpha = 0.5) +
  scale_color_viridis(option = "plasma") +
  theme_classic()

# # of species per year??
ggplot(data = FishGlob.10year.spp_survey_year.r[survey_unit == "ZAF-IND",], aes(x = year, y=spp_count_survey_year)) +
  geom_col() +
  theme_classic()

# # of hauls per year
ggplot(data = FishGlob.10year.spp_haulid_year.r[survey_unit == "ZAF-IND",], aes(x = year, y=haulid_count_survey_year)) +
  geom_col() +
  theme_classic()

-Include 4,5,6 -Consistent spatial distribution through time -Richness consistent through time -# Hauls consistent before 2001, and then also in 2005 and 2009-2010

zaf_ind_keep <- unique(FishGlob.10year.uniquehauls[survey_unit == "ZAF-IND" & month %in% c(4:6) & year %in% c(1985:2001,2005, 2009,2010), haul_id])

####Combine all lists that have _keep

#all objects with _keep
list_obj <- ls(pattern = "_keep")

#combine
fishglob_haulids_to_keep <- unlist(lapply(list_obj, get)) #243539 hauls (Started with 296799)

FishGlob.10year.spp_manualclean <- FishGlob.10year.spp[haul_id %in% fishglob_haulids_to_keep,]

#save
saveRDS(FishGlob.10year.spp_manualclean, file = here::here("data","cleaned","FishGlob.10year.spp_manualclean.rds"))

####Some surveys sample through end of year, fix these -NOTE THAT THIS NZ-CHAT SURVEY CROSSES YEAR, SO LUMP 1 and 2 with previous year